Computer systems today are able to record and store more information logs relating to their internal status or error conditions than ever before. When a computer system encounters a problem, finding the correct informational log needed to diagnose and correct the issue is an increasingly difficult task.
Packages of logs on a remote computer may be transmitted to what is interchangeably referred to herein as either a “home” computer or a central repository via various communication media, where support specialists may analyze the problem or event that prompted the transmission. The transmission is initiated by the remote computer as a transmission of log files as data packages to the central repository. Through existing log transmission methodologies, it is typical to identify and transmit common logging files that may or may not be useful to support personnel after a problem has been realized. Since these common files are sent home indiscriminately, regardless of what type of error is detected, there is no adaptation to different types of errors.
Limiting factors exist that prevent sending more logs in data packages. These factors include, for example, the connection speeds from the remote machine and the central repository and the cost of storing data files for long term analysis.
Because of these limiting factors, there are situations where the correct logs are not sent in the data package, resulting in support center's need to remotely connect to a machine or in a worse case scenario, send a person to the remote site to retrieve the required logs. This situation results in higher support costs.